sea surface
race winds,
die range of a
ìat only a
i the vicinity
:d. With a
principle at
rhe sea any-
! capital and
¡ver, enormous
: a ground-
its of ground-
in (1981) and
relopments have
le lidar for
one page
>f the Manual.
>le advances,
available but
reloped and
re review is
1NTS
>ments should,
>f view of (i)
> and coastal
resent routine
i optical and
>pectrum.
applications
) oceanographic
fraphers
i number of
)ursued in the
good; in my
i shortage of
listribution
processing and
rute the main
achieve in
1 two senses,
data. In some
real time access
image or
:rn developments
possible to
software for
software that
and is not
1 but is also
reamlining of
.cult but it
> be financed,
inding problems
>ns to the
scanners the
be made either
existing
meed Very
le two-look
ling Radio-
ied to be
.th optical
. at least in
.bution to the
wavelengths
:entage of the
ifrared wave-
i very small
id if one is
i make good
done on a
pixel-by-pixel basis, to optical scanner data then
one is going to have substantially more
spectral bands or channels in the optical region of
the electromagnetic spectrum. I would estimate that
we need at least about 20 bands well distributed
throughout the optical region of the spectrum before
we can hope to obtain reasonably reliable corrections
for atmospheric effects. Such instruments, sometimes
known as imaging spectrometers and quite probably
having more than 20 bands, are being developed with
a view to flying them in space; however, it is
likely to be quite some time before data from such
instruments will be available for us to use on a
regular basis.
As has already been mentioned, considerable
quantities of optical and infrared scanner data from
space are already available. This is supplemented
by only occasional radar data from Shuttle missions,
by ground-based radar data and by data from aircraft
for very restricted geographical areas. But there
is no regular supply of radar data from space and it
will be several years before regular supplies of
such data become available. The period of waiting
is being used (a) by practising and gaining
experience of handling microwave data using old data
from SEASAT etc., (b) by generating simulated micro-
wave satellite data, (c) by improving the algorithms
for the extraction of geophysical parameters from
microwave satellite data and (d) by trying to ensure
that data acquisition, distribution and processing
systems are properly planned and developed in
readiness for the handling of data from the oceano
graphic microwave remote sensing satellites that are
due to be launched in the late 1980s and early 1990s.
4.2 Estuaries and coastal regions
For estuarine and coastal work the situation is that,
with the new higher-resolution data from the LANDSAT
Thematic Mapper (TM) and the SPOT HRV becoming
available, work is starting on the exploration of the
opportunities provided by these new sources of data.
Essentially the new higher-resolution data opens up
new possibilities for the use of satellite data for
many smaller-scale situations that previously were
not amenable to study with satellite data. Many of
us, in the academic world at least, feel very
frustrated by the extremely high prices that are
being charged for data from these sources. In the
commercial world there is a danger that the high
cost of the data will mean that many potential
applications of the new remotely sensed data will
not come to fruition. I do not believe that this is
in the long-term interests of the builders or
operators of remote sensing satellite systems.
The spatial resolution of the new satellite data,
especially that from SPOT, is beginning to approach
that of data from aircraft-flown instruments and
this means that satellite data can begin to compete
seriously with airphotos for cartographic work
related to the land, of rivers and estuaries and of
the coastal zone. However, in spite of the increased
spatial resolution achieved, there are still serious
limitations in the way of using this high-spatial
resolution satellite data for studying phenomena
that involve rapid change. Images of any given area
are generated by LANDSAT or SPOT very much less
frequently than once per day. But there are many
estuarine or coastal problems, such as the dispersion
of a plume of industrial effluent or domestic sewage,
or the general circulation and the behaviour of
natural or man-made suspended sediment distributions,
or the mixing of fresh water and sea water, which
involve changes on a much shorter timescale than
this. They require data from a succession of passes
at frequent intervals during one complete tidal
cycle. Such data cannot be provided by a satellite
and so, for situations involving rapid change, we
can expect that data from aircraft-flown surveys
will continue to be used. One has the impression
that some people working in remote sensing have
seen aircraft only in the context of preflight test
ing of instruments to be flown in satellites or of
validation flights carried out simultaneously with
satellite passes over a given area. Some of these
people are now discovering what they ought to have
known all the time, namely that there are many
situations in which aircraft are far more appropriate
than satellites for generating remotely-sensed data.
4.3 Hydrology
Some of the comments made in section 4.2 apply also
to hydrology. To these I would add that the
possibility of serious stereoscopic image data from
space using the SPOT satellite is actively being
studied by a number of workers. While this is not
relevant to water surfaces it is, however, relevant
to the study of river basins and catchment areas.
Although the product obtained from a scanning
instrument may look superficially the same as a
photograph produced by a camera one has to be careful
when dealing with stereopairs. It is necessary to
remember that there are differences between the
principles on which scanner images and images in a
camera are generated. These differences have to be
built in to the theory that is to be used for the
extraction of topographical information from a
stereoscopic pair of images; some details are given
by Dowman (1984).
5 GENERAL CONCLUSIONS
I think my general conclusions must be that the
future looks interesting and exciting. New and
improved sources of remotely-sensed data are being
developed and made available. The scientific
remote sensing community is busy tackling many
interesting and important research topics. The
entrepreneurs are making considerable progress in
the commercial exploitation of remotely-sensed data.
One of the things at which we have to continue
working relentlessly is public relations. We must
spare no effort in the struggle to put the case for
remote sensing to politicians, to administrators
and to potential users on every possible occasion
that presents itself to us.
One of the things that I hope we shall see in the
next few years is a move towards a much greater use
of remotely-sensed data in near-real time.
Perhaps as a final remark I can come back to the
point at which I started and which I think it is
especially appropriate to make at an ISPRS meeting.
When I moved in to the field of remote sensing about
eight years ago I had the very distinct impression
that there was a great gulf between the traditional,
and very precise, activity of photogrammetry and the
activities involved in remote sensing which were
often qualitative or, at best, only semiquantitative.
In those days ISPRS was simply ISP. I detect a
considerable degree of convergence in recent years
between these two activities and I very much hope
that this will continue and be for the benefit of
everyone.
REFERENCES
Allan, T.D. (ed) 1983. Satellite microwave remote
sensing. Chichester: Ellis Horwood.
Colwell, R.N. (Editor-in-Chief) 1983. Manual of
remote sensing vols. I and II. Falls Church,
Virginia, American Society of Photogrammetry.
Dowman, I.J. 1984. Space cartography. In Remote
sensing applications in civil engineering p.
97-121. Paris, European Space Agency.
Muirhead, K. and Cracknell, A.P. 1986. Airborne
lidar bathymetry, Int. J. Remote Sensing 7:
597-614
